A friend told me that an internet company he had stayed in before had a weak ability to resist risks. Allin, the whole operation department, completely ignored the indicators such as retention and activity. The number of new users of ASO from the application market on 20 17, excluding other channels, can achieve an average of 3W per day. However, the retention rate is particularly low, and the 7-day active retention rate can only be maintained at around 10%. Later, a new product manager came to the company. The product manager saw the company's problems and gradually improved the data system of the whole company. Later, the index system of operational data gradually became clear, and the growth of company users also entered a healthy growth state, which was much more sustainable than allin's new profits at that time. He lamented that if the data analysis is good, sustainable profit growth can be achieved completely, and he deeply felt the importance of data analysis. I also totally agree with him that data analysis has great potential. Today, combined with my many years of experience in APP data analysis, I will explain some ideas of APP data analysis to you. Remember, just talk about ideas, not practice, hoping to help some partners who are interested in APP data analysis. Monitor daily data operation indicators, such as the number of downloaded users, new users, active users and paid users. It is the most basic data in operation and the core index that big bosses are most concerned about. These indicators require high accuracy and timeliness of data, so once you enter a new company or take over a new project, the first task is to sort out these data. In addition, many indicators in the operation index system are derived from these basic indicators. If the data quality of these basic indicators is not good, other derived indicators will be biased, and the biased results will be greater than the basic indicators due to the superposition of errors of multiple basic indicators. How to ensure the data quality of basic indicators? The design of user ID logic is very important. For the statistics of the number of users, the design logic of user ID directly determines the quality of data. So when you get these basic data, you should have a clear understanding of the ID logic behind the statistical data. For e-commerce and social apps, because such apps have a strong membership system, they will play a very good complementary role in accurately identifying a user. Channel analysis For an APP in the rising or falling period, the operation team will try to find as many channels as possible to attract the attention of new users. There are many channels on the Internet, including bidding channels (Baidu, sogou, App Store), seo channels (Baidu, sogou), new media channels (WeChat official account, Weibo, Tik Tok), online advertising channels (Baidu Online Alliance, Ali Mama), mobile payment channels (Today's headlines, Tencent Broadcom), free channels (QQ group, WeChat group, post bar, etc. There are many channels, so monitoring and analyzing the channel effect is very helpful to reduce the cost of obtaining customers and improve the ROI of channel promotion. Channel analysis is nothing more than monitoring the quality of each channel, which effect is better and which unit price is cheaper. Of course, we also need to monitor the follow-up performance of users in different channels and score users in each channel. We should clearly let BOSS know which channels are worth investing and which channels are rubbish. Which channels need to increase investment and which channels should choose to give up. If the operation team has sufficient resources, it can also compare and analyze the user quality between different mobile phone models, different operating systems and different regions. In short, it is to slice new users in different dimensions to monitor user performance in different dimensions. Of course, in channel analysis, there are two important issues that need urgent attention of market personnel and data analysts, that is, channel deception and channel attribution. Channel deception and channel attribution are both very complicated research topics. I will write something about these two pieces separately later, so I won't go into details here. Active users cannot satisfy all users when analyzing a product. You can't have your cake and eat it. The reason why users become active users must be that your products meet certain user needs. Studying active users helps us to improve the core function points, so their behaviors are worth studying. Therefore, active users (or core users) are the most valuable resources of the APP. Pay close attention to the dynamics of active users of the APP and listen to their voices. In the analysis of active users, we can pay attention to DAU, WAU, MAU, startup times, usage duration, DAU/WAU, DAU/MAU and other indicators. WAU and MAU reflect the total size of active users, startup times and usage duration reflect the stickiness of active users, and DAU/WAU and DAU/MAU reflect the activity of active users. In the analysis of active users, the indicators reflecting stickiness and activity deserve careful study. Take the use of time indicators as an example. This indicator is the time that users spend on the APP in a certain natural period of time. The biggest function of this indicator is to evaluate user activity and user stickiness. If the user's use time is ideal, it means that the user has high recognition and demand for the APP, and vice versa. On the other hand, think about how much time an ordinary user expects to spend every day when your application is designed. Does the user really spend the same time after going online as you expected? If the deviation is large, it means that the user's cognition of the APP is different from what you imagined at that time. At this time, you need to think about how to adjust your products to meet the user's cognition. User portrait analysis User portrait is actually the annotation of user information. Such as gender, age, mobile phone model, network model, professional income, hobbies and so on. The core work of user portrait analysis is to tag users, tag users according to the tagging rules formulated by people, read information quickly through tags, and finally extract tags to form user portraits. There are two main application scenarios of user portrait: user feature analysis and user grouping. User feature analysis is a continuous and in-depth insight into the user attributes of a specific user group, which makes the portrait of this user group gradually clear and helps enterprises understand who they are. What are the behavioral characteristics? What do you like? What are the potential needs and behavioral preferences? After understanding these characteristics, we can make a targeted analysis of the subsequent user groups. User grouping is the basis of refined operation and has been widely used in data analysis of all walks of life. For example, positioning marketing target groups to help enterprises achieve precise marketing; In order to wake up sleeping users or recall lost users, help enterprises achieve accurate push; For example, e-commerce or information app helps companies achieve personalized content recommendation and so on. Analysis of product core function transformation What is transformation? When users operate in the direction of your business value point, there is a transformation. Business value points here include but are not limited to completing registration, downloading and purchasing. In the field of analysis of Internet products and operations, transformation analysis is the most core and key scenario. Taking e-commerce website shopping as an example, a successful purchase behavior involves searching, browsing, adding a shopping cart, modifying an order, settlement, payment and other links in turn. Problems in any link may lead to the failure of users' final purchase behavior. Under the background of refined operation, how to do a good job of transformation analysis is very important. Therefore, when you want to do transformation analysis, you should think about what is the core function of your product, and then monitor the transformation rate of this core function. Different industries have different conversion rates. For example, the game APP pays more attention to the payment rate, and the e-commerce APP pays more attention to the purchase rate. Conversion rate analysis, you can also compare your products with the industry average to see where your products are in the industry. In addition, through long-term trend monitoring, we can evaluate the good and bad of different versions of the APP. User churn analysis Loss user recall is an important part of operation, and defining loss users is the starting point of user churn analysis. Lost users usually refer to those users who have used products or services, but later stopped using them for some reason. In practical work, the definition of lost users is much more complicated for products or services of different business types. For example, e-commerce products, according to the definition of user purchase behavior, how long it takes for users not to buy again is considered as user loss; For example, for content products, according to the definition of user access behavior, how long a user has not visited is considered as the loss of users; For example, video products, according to the definition of user viewing behavior, how long a user has not watched it is considered as the loss of users. Therefore, it is necessary to quantify the key behaviors of users in combination with product business types to define the lost users. User churn is a process, not a node. The agitated users will show some abnormal behavior characteristics before they officially stop using the product: the frequency of access is greatly reduced, the online time is greatly reduced, and the frequency of interaction is greatly reduced. Therefore, we need to establish an early warning mechanism of user churn through rules or machine learning modeling, predict the probability of user churn in advance, and support operations to intervene in users with high potential churn. If you have the conditions, you can compare it with the industry average, so that you can know more about the position of the loss rate of your products in the industry. In addition, we can also make portraits of lost users, which can help us better understand the characteristics of lost users. The more detailed and representative the portrait of the lost user, the higher the recall success rate. However, we know that lost users and portraits of lost users are not enough. We need to find the lost places, see where the users are lost, and then make corresponding product changes. When we clearly define the lost users, understand the portraits of the lost users, and know which channels the lost users gather in, then we must clarify the path and strategy of user recall. From the user's point of view, give users a reason to reuse products. After the recall of lost users, it is not the end. It is necessary to maintain and promote the second recall of lost users and consolidate the recall effect. User life cycle analysis What is the life cycle of APP users? It refers to the whole development process from the beginning of establishing a relationship with APP to completely separating from APP. The total value brought to APP in the whole life cycle is called life cycle value. The whole life cycle of APP users can be divided into four different periods from the perspective of user value contribution, namely, investigation period, formation period, stability period and decline period. Users in each period bring different values to the APP. (1) During the inspection, users mainly verify and inspect the functions and services provided by APP products. Once users find that products can't meet their own needs, they will quickly lose. Therefore, in product planning, we must accurately locate the needs of the target population and target users, and try to avoid the loss of a large number of users after going online. The value contribution of users in this period is low. (2) When the functions and services of the product can meet the needs of users, users will try to use the product, and the user experience of the product plays a decisive role in this process. Especially when the homogenization APP is serious, users will choose an APP with a better experience. During this period, users will really choose and decide to use products, and the value created by users will also increase rapidly. (3) Users in this stable period have the highest loyalty and activity. They will frequently use products, promote products through word of mouth, attract and recommend more users to choose products, and the user value creation in this period will reach the highest level and remain stable for a long time. (4) There are many factors that cause stable users to enter the degradation stage during the degradation period. For example, a maternal and child product, children will give up using this product when they grow up. In a word, some factors affect the satisfaction of users, which may prompt users to enter the degradation period and then leave the product completely. Once the user enters the degradation period, it is necessary to maintain the user in time. At this stage, the value created by users will decrease rapidly. Summarizing the above summarized ideas of APP data analysis is not all, such as A/B testing, heat map analysis, form analysis, path analysis and other common analysis ideas, which are not included here. With so many ideas of APP data analysis, there are actually very mature APP data analysis tools on the market, which provide us with strong analysis support. For example, there are Youmeng, MTA, Talkingdata, Ce Shen Data, Growingio, Zhuge io, Number Geeks, etc. In China, there are GA, Mixpannel, Appsee, etc. Abroad. The basic data analysis dimensions of each APP data analysis tool are almost the same, and each product has its own unique advantages. Therefore, if you want to choose a third-party data analysis tool, you should choose the appropriate data analysis tool according to your own analysis purpose and the conditions of your company. You can't always judge by your age, nor can competing products. After all, in the long river of history, there are many things that the waves in front of you are patted on the beach by the waves behind you. In the era of big data, data analysis has become the core competitiveness. As the saying goes, "know yourself and know yourself, and you will win every battle." Through professional Tik Tok data analysis, we can not only learn the latest gameplay in the industry, but also learn the popular "routines" of peers, which will get twice the result with half the effort. The big clue data platform contains the most data in the whole network, and its functions are also very comprehensive. It provides video ranking, live broadcast analysis, e-commerce data analysis and other services, which can help individuals, businesses or MCN institutions to better create and operate short videos. Most importantly, all functions are free and open to use.